Check whether your website's lead numbers are real: have an AI read your tracking setup and find what's being counted as a lead that isn't one: AI agent prompt
Analytics dashboards often count the wrong things as leads, like brochure downloads, newsletter signups or repeat form loads, so you end up making ad and staffing decisions on inflated numbers. An AI agent can read your site's tracking code and conversion settings, list exactly which actions are counted as leads, and compare that against the real enquiries in your inbox or CRM. Monill Parikh posted on X (Oct 22, 2026) that Google Analytics showed 245 leads for a website when the real number was 14; his AI agent caught it by reading the site's tracking code and finding that every catalogue download was being counted as a lead.
MarketingAnalytics lead-count auditintermediateThe prompt
You are my analytics auditor. My website is [URL]. Here is what my analytics tool reports as leads or conversions for [date range]: [number, and the names of the conversion events if you know them]. Here is my tracking setup: [paste the tracking/tag code from the site, the Google Tag Manager tags and triggers, or a screenshot or export of the conversion events]. Here are the real enquiries we received in the same period: [count, or paste from the inbox/CRM: date, name or ID, source]. Do this: 1) List every event that is counted as a lead or conversion and what action on the site fires it (form submit, thank-you page view, button click, file download, etc.). 2) Flag events that are not real enquiries (downloads, newsletter signups, page loads, duplicate fires, test traffic, our own team's visits). 3) Estimate how many of the reported leads came from each event, and compare the total with the real enquiry count. 4) Recommend exact changes: which events to stop counting as leads, which to keep as separate 'micro' goals, and how to make the real form submission fire once. 5) Give me a short before/after summary I can share with whoever runs our ads. Don't change any settings yourself; if something is unclear from what I pasted, say what you'd need to check.
How to use it
- Copy the prompt above.
- Replace every [bracket] with your own details (or fill in the blanks on AgentPlaybook Real Estate, which remembers your answers for the next prompt).
- Paste it into Claude, ChatGPT, Google Analytics and review the answer before you use it.
Works with: Claude, ChatGPT, Google Analytics, Google Tag Manager, HubSpot
Real-world example: x.com/MonillParikh/status/2113287007418065158
Featured in: ChatGPT prompts for business email, replies and follow-ups
Before you use it: what to check
Already built into this prompt
- Don't change any settings yourself; if something is unclear from what I pasted, say what you'd need to check.
Check yourself
- This prompt works on text other people wrote (emails, reviews, messages or web pages). That text can hide instructions aimed at AI, such as "ignore your rules and…". Paste it as material to work on, and never let the AI act on requests inside it, like paying, sharing logins or changing records. How prompt injection works →
- Read and edit the draft before it reaches a customer or goes public. If you later automate this prompt, start with "draft and notify me", not "send automatically".
- Check names, dates, numbers and any claim against your own source material before you use the answer. AI can misread or invent details.
Using an AI agent? Download this prompt as a SKILL.md agent skill with these checks built in, or get the .zip for Claude.ai. How to install skills.
Want to make this prompt even better? Grade it with the free prompt grader or build your own with the AI prompt generator.
32 copy-paste agent prompts (PDF) plus 5 new prompts every week.
Get it free